Foreign Substance Detection via Image Smoothing and Difference Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The computational complexity of spatial frequency analysis for detecting regions with predetermined spatial frequency spectra in images poses a significant load on detection apparatuses, particularly in applications like vehicles and imaging systems, making it challenging to efficiently detect foreign substances like dirt or water droplets on camera lenses.
Innovation Solution
A detection apparatus and imaging system that uses a controller to generate smoothed images and difference images, allowing for the detection of low-frequency regions without analyzing spatial frequency spectra, thereby reducing computational load and enabling the detection of foreign substances by comparing captured images with their smoothed versions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If spatial frequency analysis is used to detect foreign substances on the lens surface, then detection accuracy is improved, but computational complexity increases significantly
Solution Approach 1:
The patent divides the captured image into multiple blocks and performs frequency analysis on each block separately. This segmentation allows the system to detect foreign substances with high accuracy in specific regions without requiring computationally intensive full-image analysis, thus resolving the contradiction between detection accuracy and computational complexity.
Solution Approach 2:
The patent applies frequency analysis only to specific blocks where foreign substances are suspected, rather than analyzing the entire image. By performing partial analysis on selected regions, the system achieves sufficient detection accuracy while significantly reducing the overall computational burden.
2Reliability
If spatial frequency analysis is performed on the entire image, then foreign substance detection capability is improved, but processing time increases
Solution Approach 1:
The image is divided into multiple blocks for parallel processing. This segmentation enables the system to maintain comprehensive foreign substance detection capability across the entire image while reducing processing time by distributing the computational workload across multiple smaller regions that can be analyzed simultaneously or selectively.
Solution Approach 2:
The system performs frequency analysis on selected blocks rather than the entire image, achieving sufficient foreign substance detection capability with reduced processing time by focusing computational resources only on regions where foreign substances are likely to be present.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A detection apparatus includes an image acquisition interface that acquires a captured image captured by an imaging unit and a controller that generates or acquires a smoothed image yielded by smoothing the captured image. The controller compares the captured image and the smoothed image and detects a low-frequency region having a predetermined spatial frequency spectrum from the captured image.